Combined Deletion of Four Francisella novicida Acid Phosphatases Attenuates Virulence and Macrophage Vacuolar Escape

Abstract

Francisella tularensis is a facultative intracellular pathogen and the etiologic agent of tularemia. It is capable of escape from macrophage phagosomes and replicates in the host cell cytosol. Bacterial acid phosphatases are thought to play a major role in the virulence and intracellular survival of a number of intracellular pathogens. The goal of this study was to delete the four primary acid phosphatases (Acps) from Francisella novicida and examine the interactions of mutant strains with macrophages, as well as the virulence of these strains in mice. We constructed F. novicida mutants with various combinations of acp deletions and showed that loss of the four Acps (AcpA, AcpB, AcpC, and histidine acid phosphatase [Hap]) in an F. novicida strain (ΔABCH) resulted in a 90% reduction in acid phosphatase activity. The ΔABCH mutant was defective for survival/growth within human and murine macrophage cell lines and was unable to escape from phagosome vacuoles. With accumulation of Acp deletions, a progressive loss of virulence in the mouse model was observed. The ΔABCH strain was dramatically attenuated and was an effective single-dose vaccine against homologous challenge. Furthermore, both acpA and hap were induced when the bacteria were within host macrophages. Thus, the Francisella acid phosphatases cumulatively play an important role in intracellular trafficking and virulence.

Francisella tularensis is a gram-negative, facultative intracellular pathogen that causes tularemia in humans and a wide range of animals and is capable of surviving and replicating inside macrophages (11, 35, 36). An F. tularensis infection can be acquired through an arthropod vector bite, by ingestion of contaminated food or water, from infected animal carcasses, or by inhalation of as few as 10 bacteria (10). The Centers for Disease Control and Prevention has classified F. tularensis as a potential bioweapon due to its high levels of infectivity and lethality.

Three main subspecies of F. tularensis capable of causing disease in humans have been identified: F. tularensis subsp. tularensis (type A strain), F. tularensis subsp. holarctica (type B strain), and F. tularensis subsp. mediasiatica. Francisella novicida, a close relative of type A F. tularensis, is an infrequent cause of infection in humans but causes a lethal systemic infection in mice when it is inoculated by most routes (17). All Francisella subspecies exhibit more than 95% DNA identity.

There is no approved vaccine against tularemia in the United States or Europe (15). The current live vaccine strain (LVS) is an attenuated type B strain that provides diverse levels of protection against challenge with virulent type A F. tularensis depending on the route of immunization and the host. The unknown attenuation of the LVS strain may preclude its approval as a vaccine candidate by the Food and Drug Administration (16). Thus, the development of new vaccines and therapeutics is necessary.

Few virulence factors have been identified in Francisella, and the molecular events resulting in tularemia are still unclear. Several studies have indicated that the products of Francisella pathogenicity island (FPI) genes, such as IglC, IglD, IglA, PdpB, and PdpD, are important for virulence (9, 32, 33, 37). FPI genes are controlled by the global regulators MglA, SspA, and PmrA (5, 6, 19, 22, 24). Several studies have shown that in an FPI-mediated manner, Francisella escapes from the phagosome by 6 h and replicates in the cytoplasm (7, 8, 21, 32), which results in apoptosis or autophagy by 20 to 24 h postinfection (7).

Acid phosphatases are ubiquitous in nature and are found in both prokaryotes and eukaryotes. These enzymes catalyze the hydrolysis of phosphomonoesters at an acidic pH (29). In several bacterial species, acid phosphatases have been found to play a major role in virulence and to aid bacterial survival inside phagosomes (3, 4, 12, 18, 20, 26, 27, 28, 31). The published genome sequence of F. novicida revealed that this organism has five acid phosphatase genes (acpA [FTN_0090], acpB [FTN_1556], acpC [FTN_1061], hap [FTN_0954], and FTN_0022), and all of these genes except FTN_0022 are present in type A strains (14). In our previous study, we observed that AcpA functioned as a phosphatase and a lipase in F. novicida as deletion of acpA in F. novicida reduced the phosphatase and lipase activities 10- and 8-fold, respectively. The acpA mutant was defective for survival inside THP-1 and J774.1 macrophages, and electron microscopy studies revealed that its escape from macrophage phagosomes was delayed. In mice, the acpA mutant showed a defect in time to death, but the lethality at all of the doses tested was identical to that of the parental strain (100%) (21).

To further delineate the role of the Acp proteins in F. novicida virulence, four putative acp genes in F. novicida were systematically deleted. In the present study, we used nine different combinations of acid phosphatase gene deletions in F. novicida and investigated mutant phosphatase activity, intramacrophage survival, virulence in the mouse model, and intracellular trafficking within macrophages.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Bacterial strains, plasmids, and primers used in this study are listed in Table 1. F. novicida U112 was routinely cultured at 37°C on cysteine heart agar (CHA) (Hi-Media Laboratories, India) or in modified tryptic soy broth (TSB) (Difco Laboratories, Detroit, MI) containing 35 μg/ml ferric pyrophosphate and 0.1% cysteine hydrochloride. Kanamycin, tetracycline, hygromycin, and erythromycin were added to final concentrations of 25, 12.5 or 25, 100, and 25 μg/ml, respectively, to CHA or tryptic soy agar for selection of F. novicida U112 mutants. The other solid media used included CHOC II plates (Difco Laboratories, Detroit, MI). All genetic manipulations with F. novicida were performed in a class II biological safety laboratory. Escherichia coli DH5α cells were used to carry and propagate all vectors. When required for E. coli growth, Luria-Bertani medium (Difco Laboratories, Detroit, MI) was supplemented with kanamycin (50 μg/ml), tetracycline (15 μg/ml), hygromycin (200 μg/ml), or erythromycin (250 μg/ml). All antibiotics and chemicals were purchased from Sigma-Aldrich (St. Louis, MO). To obtain growth curves, F. novicida wild-type and mutant strains from 24-h CHA plates containing antibiotics were inoculated into modified TSB, as well as Chamberlain's medium, and the optical densities of the cultures were determined at 2-h intervals.

TABLE 1.

Bacterial strains, plasmids, and primers

Strain, plasmid, or primer	Description or sequence (5′-3′)	Source or reference
F. novicida strains
JSG1819	F. novicida U112	ATCC
JSG2660	JSG1819 with ΔacpA::kan	21
JSG2837	F. novicida U112 with ΔacpC::erm	This study
JSG2864	F. novicida U112 with ΔacpB::tet	This study
JSG2865	F. novicida U112 with Δhap::hyg	This study
JSG2866	F. novicida U112 with ΔacpA::kan ΔacpC::erm	This study
JSG2867	F. novicida U112 with ΔacpA::kan Δhap::hyg	This study
JSG2868	F. novicida U112 with ΔacpC::erm Δhap::hyg	This study
JSG2870	F. novicida U112 with ΔacpA::kan ΔacpB::tet ΔacpC::erm	This study
JSG2869	F. novicida U112 with ΔacpA::kan ΔacpC::erm Δhap::hyg	This study
JSG2871	F. novicida U112 with ΔacpA::kan ΔacpB::tet ΔacpC::erm Δhap::hyg	This study
JSG2872	F. novicida U112 with ΔacpA::kan ΔacpB::tet ΔacpC::erm Δhap::hyg carrying pHap	This study
JSG 2949	F. novicida U112 with ΔacpA::kan ΔacpB::tet ΔacpC::erm Δhap::hyg carrying pAcpA	This study
E. coli DH5α	F− φ80dlacZΔM15 Δ(lacZYA argF)U169 endA1 recA1 hsdR17 deoR thi-1 supE441 gyrA96 relA1	BRL
Plasmids
pUC18	High-copy-number cloning vector	New England Biolabs
pCLHyg	Plasmid carrying hygromycin cassette	30
pKK214	Low-copy-number expression vector with GroEL promoter of F. tularensis LVS	1
pIM13	Plasmid carrying erythromycin cassette	23
pAcpBUp	pUC19-acpB upstream region	This study
pAcpBUpDn	pUC19-acpB upstream and downstream regions	This study
pAcpB-Tet	pAcpBUpDN with Tet cassette	This study
pHapUp	pUC19-hap upstream region	This study
PhapDn	pUC19-hap upstream and downstream regions	This study
PhapUpDn	pHapUpDN with Hyg cassette	This study
pHap	pKK214 with hap for complementation	This study
pAcpA	pKK214 with acpA for complementation	21
Oligonucleotide primers
JG1332	CGGAATTCCTAAAATTATATGTTTCTAACAGCGTCAAAAATATTCGCACAATGC
JG1333	GGGGTACCAGAATTATTTTAGACCTAATTCCTTTAGTCTCTCGTTAAGAA ATATCTC
JG1334	AACTGCAGTTAGTATACTGATTTTGATTTAAAGATTAGAAAATTACCTA
JG1335	ACATGCATGCGTCTATTTGTTTCTATGCATTGGAATATATCA
JG866	CGGAATTCCTGCAGAAGCTTAGGAGGGTTTTTAATGAAATCTAACAATGCGCTCATCGTCATC
JG867	TTCGATCGAATCGATGGTACCTCAGGTCGAGGTGGCCCGGCTCCATGCACCG
Dn-primer	ACTACTGGGCTGCTTATGCAACACGTAGCTGCATAAGCAGTGAC
Up1-primer	GCTGCTAACAAAGCCCGAAAGGAACAAATCCAATGTATGGCAGCTGGC
Dn1-primer	GCGAATCATCATCAACCACTATGGG
Erm1	TGCATTAGGAAGCCCAGTAGT
Erm2	TTCCTTTCGGGCTTTGTTAGCTGC
JG1340	CGGAATTCCAATTCCTCTATTAAATTACTTCTTTCTCATTTGATTCCATCA
JG1341	GGGGTACCTTTGTTGTAATACCTATTATTTAATCCAATCGCGAGG
JG1342	AACTGCAGGCAGGTCCTGCGGTTGTAC
JG1343	ACATGCATGCAGGCTTTGTAAAAGCAGCTTCATTGGAG
JG1425a	GGGGTACCTGTTGTTTCAAGTTTTGATAATGATTAAAAATAATAGGAGTTAAAA ATGAAAAAGCCTGAACTCACCGCG
JG1426a	AACTGCAGCTATTCCTTTGCCCTCGGACGAGT
JG1494	GCACTGCAGAGGAGGGTTTTTAATGAAAAAAATATTCACTACCGGCATTTTAACC
JG1495	CGGAATTCGTAAAGTTGAAAAGAAGAGCAATTTTTATATTGCTG
JG1540	CGATGAGGCAAAATCAATTTACACAAG
JG1541	CTTTCCAGTGCTTACGACATACAG
JG1708	GTGGCGAAATTATATTTTAGATATTCGG
JG1709	TGCCAAACTTGCTCTTGA
JG1710	CCACAGCTCTAGAGCAG
JG1711	GGAGTATCAGTATCAGCACC

Genetic manipulations.

Plasmid DNA was isolated from E. coli using a Qiagen (Valencia, CA) miniprep kit as described by the manufacturer. F. novicida chromosomal DNA was prepared and E. coli and F. novicida transformations were performed as previously described (21). The local arrangement of the loci surrounding the F. novicida acid phosphatase genes is shown in Fig. 1. To construct pAcpC-Erm, an upstream fragment of the AcpC gene was amplified with Up-primer and Dn-primer, and a downstream fragment was amplified with Up1-primer and Dn1-primer from F. novicida chromosomal DNA. The erythromycin cassette was amplified from pIM13 (23) with Erm1 and Erm2. All three fragments were ligated together, and an overlapping PCR was performed with Up-primer and Dn1-primer. The large PCR amplicon carrying all three fragments was inserted into the pGEM-T Easy vector (Promega, Madison, WI), resulting in a plasmid designated pAcpC-erm.

FIG. 1.

Genetic arrangements of five acid phosphatase genes in the published genome sequence of F. novicida.

To construct pAcpB-Tet, upstream and downstream regions of the AcpB gene were amplified by PCR using primers JG1332 and JG1333 and primers JG1334 and JG1335, respectively, from F. novicida chromosomal DNA. A fragment containing 2 kb upstream of acpB was cloned into the EcoRI/KpnI sites of pUC19, resulting in a plasmid designated pAcpBup. Similarly, the region corresponding to 1,998 bp downstream of acpB was cloned into the PstI/SphI sites of pAcpBup, resulting in a plasmid designated pAcpBupDn. The F. tularensis outer membrane protein (YP_169847) promoter was incorporated into the tetracycline (tet) cassette forward primer JG866 and amplified from the pKK214 vector using JG867 as the reverse primer. The amplified tetracycline cassette was digested with the KpnI and PstI enzymes and ligated into the KpnI/PstI-digested and dephosphorylated pAcpBUpDn vector to obtain the final AcpB suicide vector, pAcpBUpDn-tet.

The suicide vector used for deletion of the histidine acid phosphatase (Hap) gene was constructed using JG1340 and JG1341 as the upstream-region primers and JG1342 and JG1343 as the downstream-region primers. The 1,986-bp upstream PCR product was ligated into the EcoRI/KpnI sites of pUC19, resulting in a plasmid designated pHapUp. The 1,895-bp hap downstream fragment was ligated into the PstI/SphI sites of pHapUp, resulting in a plasmid designated pHapUpDn. The hygromycin (hyg) cassette was amplified by using JG1425a and JG1426a from CLHyg (30) by incorporating the F. tularensis outer membrane protein gene (YP_169847) as the promoter into the JG1425a primer, and the final vector was designated pHapUpDn-hyg.

Plasmids pAcpBUpDn-tet, pHapUpDn-hyg, pAcpC-erm, and pAcpA-Kan (21) were moved into F. novicida by cryotransformation in different combinations to create mutants with single, double, triple, and quadruple acid phosphatase mutations in F. novicida. The strains are shown in Table 1. All the restriction endonucleases, the Klenow fragment, calf intestine phosphatases, Taq, Pfx, and T4 DNA ligase were purchased from New England Biolabs (Ipswich, MA) or Invitrogen (Carlsbad, CA) and were used as specified by the manufacturers.

For complementation studies, the hap gene was amplified by PCR using primers JG1494 and JG1495 and cloned into pKK214groEL such that hap was expressed from the groEL promoter. The resulting plasmid, pHap, and pAcpA (21) were introduced into the quadruple mutant by cryotransformation as previously described (21) and selected on plates containing 25 μg/ml tetracycline. Multiple colonies were screened to identify one colony containing the complementing plasmid. Both of the complemented strains were confirmed by using reverse transcriptase PCR and determining the Acp enzyme activity (which showed that there was a 2.1-fold increase in total acid phosphatase activity for the pAcpA-complemented strain compared with the wild-type strain and that the acid phosphatase activity was nearly identical to that of the wild type for the pHap-complemented strain), and they were subsequently examined using intramacrophage survival assays (see Fig. S2 in the supplemental material).

Acid phosphatase assay.

The F. novicida wild-type strain and acid phosphatase mutants were grown overnight at 37°C in TSB with the appropriate antibiotics, and cell pellets were collected by centrifugation at 8,000 × g for 10 min. The pellets were washed three times in phosphate-buffered saline (PBS) and reconstituted in sodium acetate buffer (pH 4.5). The cultures were normalized based on the optical density at 600 nm (OD₆₀₀) and lysed by sonication at a constant output of 50 W for a total of 180 s. Cell debris and unbroken cells were removed by centrifugation at 15,000 × g for 15 min at 4°C. The total protein content in the supernatant of each sample was determined with a Bio-Rad protein assay kit (Bio-Rad, Hercules, CA) and was normalized with sodium acetate buffer. The acid phosphatase activity in the supernatants was detected as described by Aragon et al. (2).

Quantitative real-time PCR (qRT-PCR).

The J774.1 and THP-1 cell lines were obtained from the American Type Culture Collection (Manassas, VA). These cell lines were cultured and maintained as previously described (21). In brief, ∼10⁹ THP-1 mononuclear cells were induced with phorbol myristate acetate (PMA) (10 ng/ml) at 37°C in the presence of 5% CO₂ for 48 h in 75-mm tissue culture flasks. Macrophages were washed twice with THP-1 complete medium (RPMI 1640 with 10% fetal bovine serum, 4.5 g/liter glucose, 1.5 g/liter sodium bicarbonate, 10 mM HEPES, 1 mM sodium pyruvate, and 0.05 mM β-mercaptoethanol) before infection. Macrophages were infected with F. novicida at a multiplicity of infection (MOI) of ∼100, and flasks were incubated for 2 h at 37°C. After incubation, macrophages were washed twice with RPMI 1640, the medium was replenished with complete medium containing 50 μg/ml gentamicin, and the preparations were incubated for 30 min to eliminate extracellular organisms. Then the macrophages were washed, the medium was replenished with fresh medium, and the preparations were incubated until further processing. At different time intervals, the medium was removed, 15 ml of sterile DNase- and RNase-free water was added to the flasks, and the preparations were incubated for 10 min at 37°C for osmolysis of macrophages (lysis of macrophages was confirmed by examination with an inverted microscope). Cell lysates were centrifuged at 1,500 × g for 5 min at 4°C to remove the host cell debris, and each supernatant was collected in a new centrifuge tube and centrifuged for 10 min at 10,000 × g at 4°C. The supernatant was discarded after this centrifugation, and the bacterial cell pellet was processed for RNA isolation using an RNeasy kit (Qiagen, Valencia, CA). Concomitant with these experiments, F. novicida was inoculated into 100 ml of THP-1 complete medium and incubated in a shaking incubator at 37°C for 24 h. The OD₆₀₀ of cultures were determined at different time intervals (e.g., the OD₆₀₀ were 0.28 at 2 h and 0.73 at 24 h), and 10 ml of each bacterial culture was pelleted by centrifugation at 14,000 × g for 10 min and processed further for RNA isolation. Similarly, F. novicida grown in TSB-cysteine (0.1%) was processed for RNA isolation (e.g., the OD₆₀₀ were 0.5 at 2 h and 1.13 at 24 h). The RNA quality and quantity were determined by using the Experion automated electrophoresis system (Bio-Rad, Hercules, CA). Five micrograms of total RNA was reverse transcribed to cDNA by using Superscript-II RNase H reverse transcriptase and random hexamer primers (Invitrogen, Carlsbad, CA) and normalized based on the DNA concentration. Ten nanograms of the converted cDNA was used for quantitative PCR with a SYBR green PCR master mixture in a Bio-Rad iCycler apparatus (Bio-Rad, Hercules, CA). Relative quantification was used to evaluate the expression of chosen genes. All primers were designed to give 200- to 220-nucleotide amplicons, have a G+C content of 30 to 50%, and have a melting temperature of 58 to 60°C. Relative copy numbers (RCN) and expression ratios of selected genes were normalized to the expression of housekeeping genes (16S rRNA and/or dnaK) and were calculated as described by Gavrilin et al. (15).

To examine potential polar effects of antibiotic cassettes on genes downstream of acpC and hap (FTN_1060 and FTN_955, respectively; the only two acid phosphatase genes studied with potentially cotranscribed downstream genes) (Fig. 1), we performed reverse transcriptase PCR using cDNA from F. novicida acpA::kan acpB::tet acpC::erm hap::hyg (ΔABCH) and wild-type F. novicida strains as described previously (22) (see above). This analysis revealed that there was no change in expression of the genes downstream of acpC and hap in the mutant compared with the wild-type strain (see Fig. S1 in the supplemental material).

Intramacrophage survival assays.

Intramacrophage survival assays were performed with J774.1 murine macrophages and THP-1 cells. In brief, 2 × 10⁵ J774.1 and PMA-induced THP-1 macrophages were seeded in 24-well tissue culture plates. Monolayers were infected with wild-type F. novicida and the acid phosphatase mutants at an MOI of ∼50:1. Two hours postinfection, wells were washed three times with PBS, and fresh medium containing 50 μg/ml gentamicin was added for 30 min. After gentamicin treatment, cells were washed three times with PBS and either lysed or incubated in the presence of medium containing 10 μg/ml gentamicin for an additional 22 h. Macrophages were lysed with 0.05% sodium dodecyl sulfate, and dilutions of the lysates were plated on CHOC II plates to enumerate the CFU.

Mouse virulence studies.

For vaccination and challenge studies, anesthetized female BALB/c mice (6 to 8 weeks old; Harlan Sprague) were inoculated intranasally with ∼10⁶ bacteria in 25 μl (total volume in one nostril). The 50% lethal dose (LD₅₀) of F. novicida administered intranasally has been calculated to be ∼10 CFU (19). The original inoculum was serially diluted and plated on CHOC II plates to enumerate the CFU. Mice were challenged with the F. novicida wild-type strain 35 days postinfection. Dissemination and clearance of bacteria were determined by the CFU assay using harvested liver and spleen homogenates.

Transmission electron microscopy.

Transmission electron microscopy was performed as described previously (21). Briefly, PMA-induced THP-1 macrophages were incubated with the F. novicida wild type or quadruple mutant at an MOI of 100:1 in a plastic four-well chamber slide. After 2, 6, 12, and 24 h of incubation at 37°C in the presence of 5% CO₂, the wells were washed and fixed immediately with 2.5% warm glutaraldehyde for 5 min and then with a combination of 2.5% glutaraldehyde and 1% osmium tetroxide in 0.1 M sodium cacodylate (pH 7.3) for 15 min at 4°C (8). The cells were then stained with 0.25% uranyl acetate in 0.1 M sodium acetate buffer (pH 6.3) for 45 min, processed further, and viewed by transmission electron microscopy using an FEI Technai G2 Spirit microscope at 60 kV. Multiple fields were examined for bacteria, as described by Mohapatra et al. (21), and identified bacteria were determined to be intraphagosomal or cytosolic. The criterion for being intraphagosomal was visualization of a phagosomal membrane surrounding the bacterium that was more than 50% intact.

RESULTS

Construction of an F. novicida strain with multiple acid phosphatases deleted.

We wanted to create an F. novicida strain devoid of the four acid phosphatases found in type A strains. To begin, the pAcpC-Erm plasmid was introduced into F. novicida by cryotransformation. The transformants were designated F. novicida acpC::erm (ΔC) mutants. Primers with sequences outside the acpC deletion (JG1520 and JG1521) were used to confirm the ΔC mutants by PCR (Fig. 2A). The wild-type strain produced a 2,160-bp PCR product, whereas the mutant strain produced a 2,700-bp band, a difference of 540 bp, which is the expected size difference. Similarly, the pAcpA-Kan plasmid was introduced into the F. novicida ΔC strain by cryotransformation. The transformants resulted in F. novicida acpA::kan acpC::erm (ΔAC) mutants. Primers JG996 and JG999 were used to confirm the ΔA mutant by PCR (Fig. 2B). The wild-type strain produced a 3,156-bp PCR product, whereas the mutant strain produced a 2,560-bp band, thus showing the expected 596-bp shift when the acpA gene was replaced with a Kan cassette.

FIG. 2.

PCR confirmation of the constructed F. novicida acid phosphatase mutants using primers upstream and downstream of the genes of interest. (A) PCR products resulting from amplification of F. novicida and ΔABCH mutant genomic DNA with primers JG1540 and JG1541. The acpC gene (750 bp) was replaced by an erythromycin cassette (1,290 bp). (B) PCR products resulting from amplification of F. novicida and ΔABCH mutant genomic DNA with primers JG996 and JG999. The acpA gene (1,545 bp) was replaced by a kanamycin cassette (900 bp). (C) PCR products resulting from amplification of F. novicida and ΔABCH mutant genomic DNA with primers JG1332 and JG1335. The acpB gene (585 bp) was replaced by a tetracycline cassette (915 bp). (D) PCR products resulting from amplification of F. novicida and ΔABCH mutant genomic DNA with primers JG1340 and JG1343. The hap gene (1,209 bp) was replaced by a hygromycin cassette (1,559 bp). FN, parental F. novicida strain.

To generate the triple mutant, the pAcpB-Tet plasmid was introduced into ΔAC by cryotransformation. The resultant strain was designated F. novicida acpA::kan acpB::tet acpC::erm (ΔABC). The ΔB mutant was confirmed by PCR using primers JG1332 and JG1335 (Fig. 2C). A 4,590-bp PCR product was produced by the wild-type strain, whereas the ΔABC mutant produced a 4,920-bp band, so again the fragment sizes matched the expected size difference.

The pHap-Hyg plasmid was introduced into the F. novicida acpA::kan (ΔA), ΔAcpC, ΔAC, and ΔABC strains by cryotransformation. The mutants were subsequently confirmed by PCR and sequencing; the PCR analysis was performed using primers JG1340 and JG1343 (Fig. 2D). The sizes of the wild-type and mutant strain PCR fragments were the predicted sizes. The resultant strains were designated F. novicida hap::hyg (ΔH), F. novicida acpA::kan hap::hyg (ΔAH), F. novicida acpC::erm hap::hyg (ΔCH), F. novicida acpA::kan acpC::erm hap::hyg (ΔACH), and F. novicida acpA::kan acpB::tet acpC::erm hap::hyg (ΔABCH). Most of the mutant strains grew like the parental F. novicida strain in modified TSB and Chamberlain's medium; the exceptions were the strains containing both acpC and hap deletions, whose growth lagged slightly until 16 h, when the optical densities were then equal to that of F. novicida (data not shown).

Measurement of acid phosphatase activity.

The Acp enzyme activity was expressed in relative fluorescence units (RFU) and was determined using the whole-cell lysates and difluoromethyl umbelliferyl phosphate as the substrate. The results showed that there were significant reductions in Acp activity for ΔA and ΔABCH (2,500,976 RFU [82% reduction] and 1,388,431 RFU [90% reduction], respectively) compared to wild-type F. novicida (13,894,312 RFU) (Fig. 3). Both of the triple-mutant strains (ΔACH, 5,074,931 RFU; ΔABC, 5,364,546 RFU) and the ΔCH strain (5,108,835 RFU) showed Acp activity that was more than 60% reduced compared with the wild-type activity, while the ΔAH mutant and the other single mutants showed 18 to 27% reductions in Acp activity.

FIG. 3.

Acid phosphatase activities determined using whole-cell lysates of F. novicida and six acid phosphatase mutants by a fluorimetric method with difluoromethyl umbelliferyl phosphate as the substrate. The results are expressed as percentages of the wild-type phosphatase activity, and the error bars indicate standard deviations. An asterisk indicates the P value is <0.005 for a statistical comparison of mutant and wild-type activities, as determined by Student's t test. WT, wild type.

Expression of acid phosphatase genes in THP-1 macrophages.

To understand the role of the Acps inside host cells, PMA-induced THP-1 macrophages were infected with F. novicida. At various time intervals, bacterial RNA was isolated and processed for use in qRT-PCR. The qRT-PCR results were expressed in RCN. The results indicated that within macrophages, acpA (570.8 RCN in THP-1 macrophages versus 2.6 RCN in THP-1 complete medium) and hap (655.7 RCN in THP-1 macrophages versus 61.9 RCN in THP-1 complete medium) were induced 219- and 10-fold, respectively, at 2 h postinfection and that there was a gradual reduction in transcription of about 50% over the next 22 h (Fig. 4). The expression of acpB (1.2 RCN in THP-1 macrophages versus 0.91 RCN in THP-1 complete medium) and the expression of acpC (1.5 RCN in THP-1 macrophages versus 1.38 RCN in THP-1 complete medium) were very low and showed no significant changes in medium compared with macrophages during the time course of the experiment (Fig. 4).

FIG. 4.

Differential expression of acid phosphatase genes inside THP-1 macrophages compared with F. novicida grown in THP-1 complete medium. The data indicate the change in expression between the two conditions and are the data for a representative experiment done in triplicate that was repeated twice with similar results. An asterisk indicates the P value is <0.005, as determined by Student's t test.

Intramacrophage survival assays.

The acid phosphatase mutants were examined to determine their entry into and survival within J774.1 and PMA-induced THP-1 macrophages. In both types of macrophages, the entry of all strains was similar (Fig. 5A and 5B). The ΔABCH mutant exhibited the highest level of attenuation in J774.1 cells after 24 h of infection (4-log decrease), while the ΔACH and ΔABC mutants showed a >3-log decrease in the number of CFU at 24 h postinfection. In THP-1 cells the ΔABCH, ΔACH, and ΔABC mutants showed a >1,000-fold decrease in the number of CFU at 24 h postinfection, but the level of survival of the ΔABCH strain was ∼5-fold less than that of the triple mutants over most of the time course of the experiment (Fig. 5B). All single and double mutants showed some attenuation of growth in both types of macrophages, but the attenuation was less than that of the triple or quadruple mutants (data not shown). Thus, the overall observation for both types of macrophages was that there was a progressive loss of survival and replication upon accumulation of acp mutations in a single strain. Complementation was performed with acpA and hap individually in the ΔABCH strain. The two acid phosphatases resulted in similar 2-log increases in intramacrophage survival, although there was still a 10-fold decease in survival compared with the wild-type strain (see Fig. S2 in the supplemental material). These data suggest that the acid phosphatases are directly responsible for the observed phenotype in macrophages and likely in other in vitro and in vivo assays.

FIG. 5.

(A) Intramacrophage survival assays performed with J774.1 murine macrophages infected with F. novicida (⧫) or the mutant derivative ΔACH (□), ΔABC (▵), or ΔABCH (○). (B) Intramacrophage survival assays performed with the PMA-induced THP-1 human macrophage-like cell line infected with F. novicida (⧫) or the mutant derivative ΔACH (□), ΔABC (▵), or ΔABCH (○).

Mouse virulence study.

To determine if any of the acp mutants exhibited a virulence defect in the mouse model, 6- to 8-week-old female BALB/c mice were infected with the various acid phosphatase mutants and the wild-type strain by the intranasal route. F. novicida ΔA-, ΔC-, ΔH-, ΔAC-, ΔAH-, and ΔABC-infected mice died within 2 to 10 days after infection (data not shown) at a dose of 10³ CFU. However, all the mice infected with ΔCH, ΔACH, and ΔABCH at a dose of 10³ CFU survived until 9 weeks postinfection (data not shown). When a dose of 10⁶ CFU was used, only mice infected with the ΔABCH mutant showed 100% survival at 9 weeks postinfection (Fig. 6A), although the ΔCH and ΔACH strains showed significant reductions in the time to death. A high bacterial burden in the liver and spleen correlated with mouse death in all strains tested (data not shown).

FIG. 6.

(A) Mouse virulence assays. BALB/c mice (n = 5) were anesthetized and infected using various doses of F. novicida and the acid phosphatase mutants by the intranasal route. ⧫, 10³ CFU F. novicida; ⋄, 10⁶ CFU ΔACH; □, 10⁶ CFU ΔABC; ×, ΔABCH 10⁶ CFU. (B) Bacterial loads in the livers and spleens of BALB/c mice (n = 5) challenged intranasally with F. novicida 35 days after vaccination with the ΔABCH strain (both at a dose of 10⁶ CFU). Mice were sacrificed at various time points after F. novicida infection to determine the fate of the challenge organisms.

To determine the protective capacity of the ΔABCH attenuated strain, mice vaccinated with either 10³ or 10⁶ CFU of ΔABCH were challenged intranasally with F. novicida (10⁶ CFU/mouse) 35 days after vaccination. None of the challenged mice given either dose died, and at days 5, 10, 15, and 21 postchallenge, mice from each group were sacrificed and their organ burdens were determined (Fig. 6B). The results showed that less than 10³ bacteria could be recovered from the liver and spleen 5 days postchallenge, and the numbers steadily decreased over the next 16 days (Fig. 6B) to undetectable levels (liver) or nearly undetectable levels at >21 days postchallenge. In addition, the mice vaccinated with the ΔCH and ΔACH strains at a dose of 10³ CFU were challenged intranasally with F. novicida (10³ CFU/mouse), and none of the vaccinated mice survived the challenge with this dose (data not shown). Thus, only the ΔABCH mutant showed both severe attenuation and a strong protective capacity as a singe-dose live vaccine.

Transmission electron microscopy.

To examine the uptake and trafficking of the wild-type strain and ΔABCH mutant, we observed these strains in THP-1 macrophages by transmission electron microscopy at 2, 6, 12, and 24 h postinfection. We found that at 2 h postinfection, approximately 98% of the wild type and the ΔABCH mutant were within phagosomes with intact vacuolar membranes (Fig. 7A). By 6 h postinfection, more than 50% of the wild-type bacteria had escaped from the phagosomes and replicated inside the cytoplasm. However, most of the ΔABCH mutant bacteria still remained in membranous vacuoles (Fig. 7B). At 12 h postinfection, we found few enclosed vacuoles containing wild-type bacteria as nearly all of the bacteria were in the cytoplasm. In contrast, the ΔABCH mutant bacteria were enclosed in single-layer and multilayer membrane vacuoles (Fig. 7C). By 24 h postinfection, 40% of the wild-type bacteria were within membrane vesicles (Fig. 7D), many of which had multiple membranes indicative of the autophagy pathway. The ΔABCH mutants remained within the multilayer vacuoles. Quantitative data for phagosomal residence are shown in Fig. 7E.

FIG. 7.

Transmission electron microscopy images of the PMA-induced THP-1 human macrophage-like cell line infected with F. novicida ΔABCH (left panels) and F. novicida (right panels) obtained 2 h (A), 6 h (B), 12 h (C), and 24 h (D) postinfection. An asterisk indicates the double or multilayer membrane of a vacuole indicative of autophagy, and the arrows indicate representative bacteria in the macrophages. (E) Quantitative assessment of bacteria within and outside phagosomes in a minimum of 300 cross sections per test group (≥500 bacteria scored). FN, parental strain.

DISCUSSION

F. tularensis is classified as a category A select agent by the Centers for Disease Control and Prevention due to its low infectious dose by the aerosol route and its high potential for lethality. Currently, there is no approved vaccine for tularemia, but there is a strong interest in developing an efficient vaccine based on rational design. Live attenuated vaccines are a promising approach, and data presented here demonstrate the potential of a new vaccine based on deletion of multiple Francisella acid phosphatases.

Acid phosphatases hydrolyze a wide variety of physiologically meaningful substrates, including proteins and peptides with phosphorylated tyrosines, inositol phosphates, phospholipidlike molecules, AMP, ATP, glucose and fructose 6-mono- or diphosphates, NADP, and ribose 5-phosphate (29). In our previous studies, we demonstrated that acid phosphatase A (AcpA) is a virulence factor that is required for intramacrophage survival and efficient escape of F. novicida from the macrophage phagosome (21). AcpA (57 kDa) is a polyspecific acid phosphatase that shows no significant global amino acid sequence similarity with any protein in the Protein Data Bank, and it has been shown to inhibit the neutrophil respiratory burst (26). Structural analysis of AcpA showed that it is more similar to alkaline phosphatases with a serine nucleophile and coordinated metal ion center (13). This protein also possesses phospholipase C-like activity. Besides the AcpA gene, four additional genes encoding proteins showing homology to acid phosphatases (AcpB [FTN1556], AcpC [FTN1061], Hap [FTN0954], and an unnamed Hap homolog [FTN0022]) are present in the genome sequence of F. novicida (14). Genomic DNA for FTN0022 is not present in type A strains, and FTN0954 may or may not be functional in type A strains as part of the gene has been deleted. The analysis of Francisella spp. histidine acid phosphatase (Hap [FTN0954]) suggested that it is a 37.2-kDa phosphatase, shares 41% identity with the Legionella pneumophila major acid phosphatase, and contains a classical acid phosphatase RHGXRXP motif (13). AcpB and AcpC have not been structurally analyzed yet.

In the present study, we hypothesized that a strain with multiple acid phosphatases deleted would be unable to escape from the phagosome and would be attenuated for virulence. Thus, nine different acp mutant strains were constructed, culminating in a strain with the four primary acp genes deleted (the genes present in type A strains). Based on broth culture and measurement of the acid phosphatase activities of the various strains, it appeared that AcpA is the primary protein contributing to the overall acid phosphatase activity of the bacterium. The quadruple mutant exhibited 10% of the wild-type level of Acp activity, which may have been due to the remaining activity of the putative F. novicida-specific Hap-type phosphatase (FTN0022). There was variability in the contribution of AcpA to the overall acid phosphatase activity as the ΔACH and ΔABC strains had more activity than the ΔA strain. While the reason for this is unknown, it may be due to compensatory mutations or to transcriptional regulation increasing the activity of AcpB (in the ΔACH mutant) or Hap (in the ΔABC mutant). We tested all nine acp mutants for intramacrophage survival in J774.1 and THP-1 macrophages at 2, 12, and 24 h postinfection. The data show that the mutants were deficient in survival and replication in host macrophages and that the defect became more evident upon accumulation of the deletions in a strain. This result was likely due to the ability of the Acps to functionally complement one another. While we were unable to fully complement the ΔABCH mutant with each deleted gene due to technical issues, introduction of either acpA or hap on a plasmid resulted in a marked increase in intramacrophage survival of the ΔABCH strain, but not to wild-type levels of survival. Thus, the phenotypes of the ΔABCH strain appear to be directly related to the functions of the deleted acid phosphatases.

It is clear from electron microscopy studies that intramacrophage growth of the acp mutants correlates with the ability to escape from the vacuole and enter the cytoplasm, as the quadruple mutant exhibits more severe attenuation than the acpA mutant in macrophage survival and in phagosomal escape (21). Similar to observations by Checroun et al. (7), the wild-type strain became associated with vacuoles again after 12 h postinfection. The vacuoles had multilayer membranes, suggestive of autophagy vacuoles. The ΔABCH strain first associated with these vacuoles at the 12-h time point, and there was increased association at 24 h postinfection. It is not known why the acp mutants do not escape from the phagosome or why they do not replicate in this location, although the lack of growth is presumably due to productive phagolysosomal fusion and/or NADPH oxidase assembly and function. These questions will be explored in subsequent studies.

To begin to understand the role of Acps inside host cells, F. novicida RNA was isolated from the THP-1 infected cells at different time points after infection. Simultaneously, RNA was isolated from F. novicida growing in vitro in TSB with 1% cysteine, as well as in THP-1 complete medium, and qRT-PCR was performed for all of the acid phosphatase genes. The data show that the expression of acpA (219-fold) and the expression of hap (10-fold) were induced in vivo, while the expression of acpB and the expression of acpC were not induced. This suggests that regulatory factors responsive to the intracellular environment activate acpA and hap; however, these regulators are not known yet. Preliminary microarray experiments did not show that PmrA or MglA affects expression of these genes upon growth in vitro (data not shown).

At a dose that was 3 logs above the LD₅₀ for F. novicida U112, only the ΔCH, ΔACH, and ΔABCH strains were attenuated, while only the ΔABCH strain was completely attenuated at a dose that was 5 logs above the LD₅₀. The quadruple mutant was defective for growth and survival inside macrophages and was unable to escape from the phagosome, which appear to be common characteristics of avirulence in Francisella (34). Interestingly, although mice infected with the quadruple mutant eliminated the bacteria in the spleen and liver within 1 week after infection (data not shown), all of the ΔABCH-vaccinated mice survived upon challenge with wild-type F. novicida (10⁶ CFU). Additionally, while several Francisella mutants (such as iglC and mglA mutants) that are defective in escape from the phagosome show partial or limited protection in the mouse model (25; K. E. Klose and J. S. Gunn, unpublished data), the ΔABCH strain appears to be an exception to this rule. Thus, this quadruple mutant warrants further investigation as a live attenuated vaccine against tularemia.